A method for obtaining consensus classifications and identifications by combining data from different experiments